Anil G wrote: > > Hi, > > Hope the question is not too simple. I am trying to make sense of > a paper that has given a pre-intervention value and the degree of > change following intervention (expressed as a percentage difference) I > am trying to figure out if the change is significant. Specifically: > > Pre intervention value is 120+/-41. After intervention, the author > reports a change of 98+/-41%. n is 11. The footnote mentions all > figures are mean +/- SE. Is it possible to recreate the post > intervention mean and SE from this so I could use a t-test? Are there > other ways to say this is statistically significant or not?
Yes. 98/41 = the number of standard errors of the difference = t (assuming the authors reported and calculated it correctly and that you interpreted their figures correctly. So, t(10) = 2.39, p = .038. ... alternatively you could plot the CI for the difference and show is doesn't include zero. Thom . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
